Physical Review X

@physrevx.aps.org

115 Followers 5 Following

@apsphysics.bsky.social's open access journal for cutting-edge research across physics and adjacent fields

Two line graphs showing mean-square generalization error on y-axis versus training data (n/d²) on x-axis for a two-layer neural network with ReLU (top) in blue and tanh (bottom) in red. Three methods are compared: a Bayes-optimal neural network (points plotted as circles), a random feature model (crosses), and GAMP-RIE (triangles). The Bayesian network achieves lower error than the random feature model and GAMP-RIE.
On the top is a scientific illustration showing how the third-order nonlinear Hall effect is measured experimentally as a third-harmonic transverse voltage V3ω driven by a low-frequency current Iω, ω ∈10-1000 Hz, or calculated theoretically as a current density J3ω driven by an electric field Eω. Below is a schematic illustration of the wavefunction, ψ, manifold in momentum, k, space, the Berry curvature and quantum metric are the imaginary, Im, and real, Re, parts of the quantum geometric tensor, respectively.
Multi-panel figure showing the formation of large scale 2D flows in rotating 3D turbulence. Top panel shows a time series of total energy, with an increase followed by steady state. Middle panel shows a log-log plot of condensate energy on the y-axis versus Rossby number, Ro, on the x-axis for multiple Reynolds numbers, shown as lines of different shades of blue, purple, pink, and orange. Bottom panels show flow visualizations of vertical vorticity as red and blue gradated images at various values of Ro: 0.0058, 0.011, and 0.5.
Figure showing the experiment. A beam of YbOH molecules is directed through a series of lasers. There are sections representing pumping, the first Ramsey pulse, free evolution, the second Ramsey pulse, and the readout. An icon shows how state populations are determined by state-selective laser-induced fluorescence detected on a photomultiplier tube.
On the left is the extracted spin state of the 𝑥1 node, 𝜎1 on the y-axis, shown throughout a small range of the simulation time on the x-axis. It depicts how, once the analog Floquet solver switches from operating in the period doubling regime and begins to exploit its nodes’ Floquet states, 𝜎1 regularly flips its spin state. On the right is the extracted spin states, 𝜎𝑖 on the y0axis, of all the nodes with time on the x0axis. It shows how the system’s spins collectively undergo periodic changes of state.
At the top left, panel (a) displays an illustration of the experimental setup involving dual-gated twisted MoTe-2 devices. The probe pulse connects to the pump, which contacts the Gr on top of a layer of hBN that surrounds the twisted MoTe-2. Panel (b) displays a schematic of an exciton in the K-valley bound to a carrier in the polarized K+ valley. This leads to an attractive polaron transition excited by σ-light. Panels (c), (d), (e), (f), (g), and (h) each display spectrally resolved transient reflectance maps for a twisted MoTe-2 device as a function of moiré filling factor, v, on the x-axis. Panels (c)-(e) show maps for a magnetic field of μ-0H = 0, while panels (f)-(h) show maps for a magnetic field of μ-0H = -6. Red dashed lines indicate various filling factors of v.
A schematic shows the structure of the superconducting thin film. Two rows each of four gray dipyramids are attached at the vertices with each vertex containing a red sphere. In the spaces surrounding the dipyramids are tricolored spheres containing equal parts green, yellow, and blue, which represent La, Pr, and Ca, respectively. A squiggly black arrow labeled h𝜈 points to one of the dipyramids, where an inset expands to show a cross section of a yellow sphere labeled ɑ surrounded by a green, plus-shaped structure labeled β, which in turn is surrounded by four red spheric sections labeled Ɣ.
Two Feynman diagrams are depicted. The top diagram shows two red wavy lines labeled "g\mu" entering and exiting black dots connected by a blue solid loop line labeled "\phi_{bra}". The bottom diagram has a dashed blue loop line labeled "\phi_{ket}" connecting similar features. A plus sign separates the diagrams.
Venn diagram showing different textile classifications. The textiles in the center satisfy all three conditions detailed in the paper. These are jersey knit on the left and “fake” knit on the right. A blue square identifies a cell unit, highlighting the difference between these two types of knits. Multiple other textiles are shown in other areas of the Venn diagram. These are non-woven textiles, chain mail, weaving, and several unlabeled textiles.
Figure with three parts to illustrate steady-state entanglement within the coherent quantum absorber scheme. The first part is a graph showing the measured concurrence on the Y axis and the Rabi drive amplitude on the X axis. The middle part shows the measured concurrence as a function of drive duration, shown by time in the X axis. The right part shows the measured expectation as values solid green bars of all two-qubit correlators for the steady state.
Figure showing the entanglement distribution scheme. It shows a Josephson parametric converter generating a two-mode squeezed state that propagates in a cascaded configuration to two distant qubits, with one at the bottom and one at the top of the schematic. Driven by the correlated fields, the qubits are shown relaxing into a state tending to the maximally entangled Bell state.
An illustration showing the top view of the 9.43° twisted V0V3 hexagonal boron nitride bilayer with a surface defect concentration of about 1%. Boron and nitrogen atoms are represented by pink and blue spheres, respectively. Two boron vacancies and one nitrogen vacancy are marked by red and blue open circles and highlighted with a dashed circle in the center of the image.
Abstract image featuring the text "Physical Review X" in the top-left corner and the white "APS" logo in the bottom-right. The background depicts a dynamic cosmic or particle explosion with bright blue and purple streaks of light that form swirling orbital loops.
On the left are five icons representing the five keystones of quantum advantage. On the top row and far left is predictability, depicted as a crystal ball with a magnifying glass that reveals text that reads “Q.E.D.” On the top row in the middle is typicality, depicted as a normal distribution. On the top row and far right is robustness, depicted as a shield with two arrows bouncing off of it. On the bottom row and far left is verifiability, depicted as a clip board with a check mark. On the bottom row and far right is usefulness, depicted as an icon of a wrench. On the right are four icons representing the realms of quantum advantage. The top left shows computation, depicted as an hour glass on a computer screen. The top right shows learning/sensing, depicted as a robot. The bottom left shows crypto./comm./strategic, depicted as two icons of people with a lock in between them. The bottom right shows space, depicted as a 1 terabyte hard drive.
Cover of the June 2026 issue of PRX titled Fifteen Years Across Physics. Against a dark, star-filled space background, a glowing blue ring surrounds a large white "15" at the center, marking the journal's 15th anniversary. Five illustrated research topics are arranged around the ring: a colorful visualization labeled "Diffusive hydrodynamics" at upper left; a 3D lattice of black spheres labeled "Gravitational waves" at upper center; layered blue panels with red particles and upward arrows labeled "Fractional quantum hall effect" at right; intersecting blue and gold light beams labeled "Neutral atom arrays" at lower left; and a cluster of blue and gold spheres labeled "Active colloids" at bottom center. The APS logo appears in the lower-left corner.
A series of diagrams illustrates the general workflow for the SciExplorer agent, proceeding from left to right. At the left, a series of text prompts gives examples of tasks assigned to the agent, which are applied to specific physics scenarios such as mechanics, waves, or quantum. To the right, a circular diagram represents SciExplorer’s reasoning workflow, which proceeds through observations, analysis, hypothesis generation, and planning and selection of an experiment. The agent then selects a numerical experiment, represented as several graphs and a plot in an orange rectangle. To the right, some lines of text represent the coding and analysis tools that SciExplorer uses to analyze its observations. The rightmost diagram portrays the last step, where the agent produces a textual summary of the reasoning chain for the human user, as well as code that can be run and tested.
A large triangle containing a color code and various qubits in black and white, presenting a 2D model equivalent to the gauged model in the paper. The qubits are grouped into three main groups labeled “W”, “X”, and “Z”. The triangle is separated into various brick segments colored either yellow, blue, green, or red, with the color code encoding a single logical qubit. Each brick is associated with a decorated Clifford stabilizer, while the intersections of top and bottom bricks are associated with “faces” corresponding to Pauli-Z type stabilizers. The left boundary is blue, the right boundary is red, and the bottom boundary is yellow.
Graphic showing two superimposed max-intensity projections of confocal stacks. Blue dots show before deformation and orange dots show after deformation. The top graphic shows simple shear, which forms a straight horizontal line. The bottom shows a line force, which forms a red bump atop a blue horizontal line.
Figure showing the experimental setup. There is a 40 Ca+ and 44 Ca+ ion trapped in a linear rf Paul trap shown by four long yellow blocks. There are two axial motional modes shown by blue cones. A red arrow pointing to one of these shows how the experiment is controlled by a 729 nanometer laser light.
Figure showing scaling laws observed in time correlation functions. The first shows the mean squared displacement as a function of time divided by the Fickian time. The second shows the self intermediate scattering function. The third shows dynamic susceptibility as a function of scaled frequency at different temperatures. The fourth shows the Fickian time compared to T-alpha above and below the mode-coupling temperature. The last shows a non-Gaussian parameter as a function of time for various temperatures.
Images of tin-vacancy centers in diamond photonic crystal cavities. At the top is an illustration of the device, showing from left the strong mirror, defect region, weak mirror, and taper end. Below, an inset image shows a simulated electric field of the fundamental cavity mode normalized to the maximum u(r). Below is a simplified level structure of the combined emitter-cavity system. The emitter is described by a transition frequency ωa between the ground state |g⟩ and the excited state |e⟩. The spontaneous emission rate is γ0 and γdep is the dephasing rate. |c⟩ denotes a photon in the cavity mode and the emitter in the ground state. The coupling strength between the cavity and the emitter is g. In the middle and to the right is a scanning electron microscope image of a free-standing photonic crystal cavity taken at a 0◦ angle colored to match the schematic. Below is a scanning electron microscope image of the diamond sample taken at a 65◦ angle showing multiple cavity arrays.
Two-panel schematic showing two regimes of BosonSampling. (a) shows a blue interferometer labeled U with many input and output ports; a few red photons enter selected inputs, and several detectors show single blue-photon outputs. A small row-and-column submatrix of U is illustrated. (b) shows more inputs contain red photons, and multiple blue photons appear at the same detectors, indicating collisions; the depicted submatrix is larger with repeated rows.
Panel B displays a series of four three dimensional graphs plotting the progression of speed minima in the neural activities of full-rank recurrent neural networks (RNNs) performing a DCD task. PC1 is on the x-axis, PC3 is on the y-axis, and PC2 is on the z-axis. The four graphs correlate to Epoch 11, Epoch 201, Epoch 1001, and Epoch 3001, from left to right. The speed minimum is denoted with diamonds. Trial type is denoted with F for fixed-point, S for slow-point, and G for ghost. Cue 1 is denoted with a blue line, and Cue 2 with an orange line. Panel D displays similar graphs but for an RNN performing an DMTS task. Here, match is colored with blue and orange lines, and no-match is colored with green and red lines. On the right, a color scale denotes the normalized speed of the fixed, slow, or ghost points increasing from dark blue to dark red.
An illustration showing the reinterpretation of an infinite-time, process tensor of Gaussian bosons as a shallow quantum circuit by taking a finite memory time approximation. The standard tensor network algorithm, abbreviated as iTEBD, produces a controlled approximation of the state computed by such a circuit as a matrix product states tensor network.
A bar chart displaying the correlations in a two-leg ladder involving 64 sites and 20 atoms with 2 cycles driving, with correlation order on the x-axis and mean |Gmc| on the y-axis. Each point of data correlates with two bars: an orange bar representing thermalized phases and a white-lined blue bar representing the MBL phases. As the correlation order progresses from 2 to 14, the mean |Gmc| steadily drops for both thermalized and MBL phases, with the MBL phases dropping more quickly. Above the bar chart, a 32x2 grid of squares visualizes the 2-leg ladder, with 20 squares in the bottom row colored red.
Graph with energy from 0 to 2.5 Joules on the y-axis and k, or wavenumber, from -𝜋 to 𝜋 on the x-axis. A yellow sine wave outlined in red with minimum energy at k = 0 shows the dynamical spin structure factor. A dashed blue line within it shows the magnon dispersion.